# WebWalkerQA (main) task ids looked at while developing jevcrawl (2026-10-12): the 30-task
# dev set, the first 10-task test set (inspected, so no longer clean) and the 10 fresh tasks
# already scored. Leave them out of accuracy runs: bench/run.py --exclude bench/tasks/seen-ids.txt
wwqa-main-105
wwqa-main-118
wwqa-main-129
wwqa-main-146
wwqa-main-18
wwqa-main-188
wwqa-main-19
wwqa-main-203
wwqa-main-22
wwqa-main-226
wwqa-main-24
wwqa-main-25
wwqa-main-250
wwqa-main-251
wwqa-main-262
wwqa-main-267
wwqa-main-298
wwqa-main-307
wwqa-main-31
wwqa-main-311
wwqa-main-341
wwqa-main-344
wwqa-main-369
wwqa-main-371
wwqa-main-383
wwqa-main-391
wwqa-main-399
wwqa-main-4
wwqa-main-406
wwqa-main-426
wwqa-main-436
wwqa-main-439
wwqa-main-447
wwqa-main-451
wwqa-main-455
wwqa-main-458
wwqa-main-474
wwqa-main-482
wwqa-main-486
wwqa-main-523
wwqa-main-53
wwqa-main-551
wwqa-main-571
wwqa-main-593
wwqa-main-641
wwqa-main-646
wwqa-main-7
wwqa-main-84
wwqa-main-9
wwqa-main-96
